by Adam T. Sutton, Senior Reporter
Landing page optimization is not an easy topic to discuss, especially with reluctant colleagues. You can explain custom landing pages, A/B tests and increased conversion rates, and many will reply with a stifled yawn.
But you care about optimization. You know it can improve results. And if you've encountered resistance, you're not alone. The internal politics of optimization could make a senator's head spin.
"When you are selling [optimization] internally, it is surprisingly complex, and there is a surprising amount of resistance to it," says Anna Talerico, EVP, Co-Founder, Ion Interactive.
We spoke with three experts to get their advice on breaking the roadblocks that stop your optimization program from rolling. Below, you will see how to begin tests without setting yourself up for disaster. We spoke with Talerico, as well as Jon Correll, CEO, Co-Founder, Conversion Voodoo, and Adam Justis, Senior Manager, Product Marketing, Adobe Systems.
Here are the tactics:
Tactic #1. Make sure you can do this
Effective optimization is not a one-off tactic. It's a process that requires steady commitment. You need to gather resources before you champion a program.
Here are three key resources you'll need:
Money
There is no getting around it, marketing budgets are tight. If you want to squeeze a testing program into your company, you need to be certain you have the money to cover a pilot program.
The exact amount of money will vary considerably with your situation. Some marketers can leverage free testing tools and internal resources to create tests and get them running. Others will have to pay for development or consultants.
Most marketers do not have "optimization" or "experimental" line items in their budgets. But the money will have to come from somewhere. Try allocating dollars from a related category, such as Web development or media spending.
If spending from the media budget, explain that your goal is to increase conversion rates, which will make your media dollars more efficient. If all else fails, try going to different departments and collecting small contributions from their budgets.
Time
Starting a testing program requires a significant amount of time. You will need to:
- Launch testing technology
- Plan tests
- Create treatments (new buttons, etc.)
- Alter webpages
- Manage tests
- Analyze the results
- And more …
"One of the biggest mistakes people make is thinking [testing] will take less time than it actually does. It is just a time-intensive process," Talerico says.
Testing expertise
You may be able to launch a free testing tool on your website without much trouble, but you need an expert to select good tests, design treatments, and analyze results. You need someone who understands optimization and testing to guide this process.
"This tool is only going to be as good as the strategy that is applied to it and the resources that are put into it," Talerico says.
If you have this expertise, more power to you. If you do not, consider reading this article as your first step in a deep research project. Read industry research, handbooks and benchmark reports. Attend conferences and workshops. Pepper experts with questions. And of course, you can hire a consultant to guide your program and fill gaps in your expertise.
Tactic #2. Prepare to be persuasive
You will not be able to launch and maintain a successful program on your own. You will have to convince coworkers to offer support. Before you visit their desks, make sure you take a look at these tips:
Do not talk about technology
"You need to be prepared to speak the language of the person that you are trying to influence. And more often than not, they are less interested in bounce rates and conversion rates. They are more interested in the bottom line," Justis says. This is doubly true if the person is an executive.
Stifle your inner nerd and avoid describing technical details unless asked.
"Never focus on the technology! Never!" Correll says. "And I'm a technology person. I started programming when I was 12 years old. I have hit my head against a brick wall a billion times, and I've slowly learned over the decades, don't ever focus on the technology. You focus on the solution. You focus on the benefit it provides."
Remember the benefits should be described in the context of revenue, not "form completions" or even "conversion rates." Specifically describe how your program will improve the bottom line.
Estimate the potential
You will inevitably be asked to estimate the potential impact of a testing program. This is a tricky question because the answer depends on many factors. You need to set expectations high enough to generate interest, but keep them low enough to prevent disappointment.
Put some numbers on the table. If someone in your organization has experience in optimization, then ask them to look at specific webpages and estimate potential lifts. Then directly correlate those lifts with revenue. For example:
- If you could get 3% more customers to complete the final step in your shopping cart process, what would that mean for your bottom line?
- "If you had 10% more leads, does it mean that you have five more opportunities every month, and that makes one more customer every month? Or does it mean you have 20 more customers a month? You have to extrapolate it out. … If you think you can get more leads in this program, what does that mean?" Talerico says. "Don't assume people are filling in the blanks. You have to really tell the story all the way to the true business impact."
Optimization case studies can also help you add estimates to your proposal (more on this in a moment).
Channel your passion
If you're going to champion this project, then you have to get people excited about it. This can be challenging, as it is easy to get bogged down by technical details.
Focus on why you care about this project -- why do you really want the company to pursue optimization? Why should anyone care about landing page tests? If you really believe your idea will drive better results, then use that conviction to inject energy into conversations.
Tactic #3. Find an IT partner
Technology powers website testing. Once you're confident enough to reach out to colleagues, the IT department should be one of the first offices you visit.
Launching and maintaining the technology will be a significant project, so you need to get an IT leader excited about the potential. Having their support will add plausibility to your pitch to executives and give you a motivated partner in a department that is typically saturated with projects.
Forge internal partnerships
Joan Renner, Senior Digital Marketing Manager, IBM, was luckier than many marketers. As manager of IBM's homepage and some of the company's content initiatives, she was approached by a member of IBM's software team who wanted to launch an optimization program.
"What he was trying to do was implement it at the corporate level so then it could go filter down to all the business units, versus having a contract over here, and one over here, and one over here," she says.
The member of the software team had the technical expertise that Renner's team needed to begin A/B testing its landing pages. That expertise was vital, Renner says, to getting the program started.
Remember: do not focus on the technology unless you're asked, even with IT managers
"If you talk about the technology, IT departments just see that as pain, right? They're going to go, 'oh my gosh! This is going to be painful! It's going to take three months! And we've got other projects, and we have this other strategy.' If you talk about it purely from the bottom line and the benefit to the company, then it's a strong argument," Correll says.
Tactic #4. Plan a pilot project
The best way to broaden support for your idea is to grab a quick success. You can do this with a well-designed pilot project. The pilot serves as proof that optimization can work, and it also introduces your organization to the testing process before investing further.
You may need to pitch executives for the resources to launch a pilot. You will of course need to be ready for their questions. Make sure you can provide information on:
- Testing technology and methodology
- Resources needed and their sources
- Potential landing page tests
- Potential benefits
- Potential drawbacks
Pick a page with impact
You want a pilot to send a quick, loud message that testing improves results, and you want it to consume few resources. The experts we interviewed emphasized starting small. Select a single page, make sure it is connected with the bottom line, and make sure there is room for improvement.
"Instead of saying 'we are going to work on our entire website,' say 'we are going to work on our checkout page where people put in their credit card,'" Correll says. "Focus on one page."
For example, an e-commerce marketer could test changes to a confusing payment page in a checkout process. A lead-nurturing marketer can focus on increasing completions on a top lead-gen form. Or you can look to your website analytics to find major drop-off points for traffic.
"[In PPC], you can start where you have a high cost-per-acquisition and drive that down. Or you can start with traffic sources driving a lot of traffic, but not a lot of conversions," Talerico says.
Do not destroy this page
If you have never tried optimization testing, then you can likely improve results on this page with simple A/B tests on single elements, such as the headline, image, button, etc. You want to avoid launching a completely redesigned page for this pilot. That approach can do more harm than good for inexperienced teams.
"You don't necessarily go to that page, wipe the slate, and send 50% of your traffic to a brand new page … You want to go to a page where there is a significant opportunity for impact, but you want to be somewhat judicious about how you treat that environment," Justis says.
Tactic #5. Present your case to executives
Be sure to complete your homework before you pitch executives. You want to outline your idea, ask for the resources for a pilot, and set very clear expectations. Here are a few tips:
Bring case studies
Search online for website optimization and testing case studies. Vendors, agencies and consultants often publish them, and our sister company, MarketingExperiments, has a library bursting with them (visit the MarketingExperiments archive using the link at the end of this article).
Look for case studies that illustrate companies in your industry (or a related industry) running effective optimization tests and improving results. Make sure the case studies show the steps taken and results achieved in numbers. If you can find an example from a competitor, that's even better.
Sell the house, not the tools
Executives are the least likely to be interested in technical details, jargon-heavy concepts or obscure metrics. They want to know how this program will impact revenue, not how you're going to build it.
"If you're going to sell a house, you sell the house," Correll says. "You don't say 'here's the drywall we used, and the nails, and the construction crew. And here, we're using these drills!'… Focus on the win. Focus on the benefit."
Set clear expectations
One often overlooked fact of optimization is that not every test improves results. Sometimes the treatment (a new button, for example) does not outperform the control (the older button). This is why you must not guarantee immediate success. Optimization is a learning process. Good tests and bad tests teach you about the audience and help you design better tests in the future.
With this in mind, be sure to clearly explain the time it will take to setup your testing technology (potentially several months) and also the time it will take to achieve real results (potentially several more months). Not only do you need the resources to make this possible, you need patience from your superiors.
Talk in the context of another project
Justis also suggested presenting your pilot in the context of a larger project. For example, you may be planning a website redesign.
"You can test different aspects of this redesign to identify what is really going to resonate with customers as opposed to putting all your eggs in one basket and hoping that what you are presenting to customers is better than what you already had."
Tactic #6. Spread your success far and wide
If your pilot is a smashing success, do not be shy. You want people to know the results so they will want to help you expand.
"The best thing you can do is tell everybody and anybody that you can who will listen," Talerico says. "It is not just about telling your boss, but your boss, and your boss's boss, and your colleague, and your friend's boss. You need to be making sure what you are doing is really visible."
Highlight the best tests
At IBM, the team convinced executives to invest more in optimization testing by presenting the results from five of its tests, just as if the tests were case studies.
"When you take those four or five case studies and present them in a way that drove positive movement, you can't help but want [to move forward]," Renner says.
Keep it small
Your next step will be to expand your pilot program, but do it slowly. You don't want a boondoggle on your hands. Instead, continue running one or two tests at a time and focus on gradual improvement.
"The biggest mistake would be to go head and multiply [the pilot] by 20 … because you're probably going to fail," Correll says.
Tactic #7. Ask for help when you need it
An effective testing process rarely fails over the long term. Sure, not every test brings a lift, but with enough expertise and a steady commitment, you will learn from every test and gradually improve results.
If you find yourself without a tangible lift after five or six tests and your pilot deadline is looming, then it is time to bring in an expert.
Don't let this bruise your ego. You might have a problem in technology, methodology, test design, results interpretation or a slew of other areas. Talk with an agency, vendor or consultant, and solicit their help. Make sure they can clearly identify and explain the problem you are having.
Useful links related to this article
MarketingExperiments - Research Directory
Landing Page Optimization: 5 questions every marketer should ask before choosing a testing tool Landing Page Test: Why less equaled (54%) more when reducing friction and highlighting value propositionChart: Best website design, management and optimization tactics for 2011 ion interactiveAdobe Test&TargetConversion Voodoo
